Requirements specification is a crucial step in software development that materializes how users interact with a system and identifies its requirements. The most common approaches are natural language and textual templates such as use cases and user stories. However, both are often limited to exemplified views of real-world system use, the so-called “normal use case” overlooking important situational and contextual details regarding users' purposeful interactions with the system. Additionally, ensuring the traceability of requirements throughout various stages of software development can be challenging. This study proposes an alternative method to address these limitations by incorporating essential elements from both approaches from an ecological perspective. The method incorporates affordances and constraints to generate more effective requirements specification and diverse scenario generation throughout the software development process. To facilitate the process, the method utilizes prompt engineering techniques, ChatGPT API and Jira API and derives an automated scenario generation system.